CURRENT FROLBLEMS IN RESEARCH
A
Fluted Projectile Points:
Their Age and Dispersion
Stratigraphicallycontrolled radiocarbondating
provides new evidence on peopling of the New World.
Problems of age, classification, and
dispersion of fluted projectile points in
eastern North America have been ex-
tensively discussed in a recent review
by Mason (1). The results of geo-
logical and archeological investigations
in the High Plains and the Southwest
are in accord with most of the inter-
pretations presented in Mason's study.
With the possible exception of Sandia
points, the earliest projectile points in
any stratigraphic context are Clovis.
These are characteristic of the Llano
complex-artifacts of a people who
specialized in hunting mammoths.
Some archeologists interpret variations
between Clovis fluted points of the
High Plains and Southwest and fluted
points of the eastern United States as
indication of temporal as well as typo-
logical differences (2), and argue that
the eastern specimens are ancestral to
Clovis points, because of the latter's
late position in an inferred develop-
mental sequence. Other archeologists
recognize minor differences but do not
consider Clovis points typologically
different from eastern fluted points
such as Enterline, Bull Brook, and
some points from the Great Lakes re-
gion (1, 3); they argue that, without
supporting geochronological evidence,
temporal differentiation is not justified.
They therefore use the name Clovis
to include related eastern projectile
points (1). Close inspection of Clovis
point collections from the Dent (Col-
orado), Miami (Texas), Blackwater
No. 1 (New Mexico), Naco (Ari-
zona), Borax Lake (California), and
Williamson (Virginia) sites reveals
that variations within individual collec-
The author is a member of the staff of the
Geochronology Laboratories, University of
Arizona, Tucson.
1408
C. Vance Haynes, Jr.
tions are as great as variations between
collections, and most of the individual
Clovis points clearly display a multiple-
fluting technique (see cover) similar to
that described for fluted points from
Michigan (4) and from the Shoop
(Pennsylvania) and Bull Brook (Mas-
sachusetts) sites (5). In the light of
these findings and of results of radio-
carbon dating interpreted in terms of
stratigraphy,I examine, in this article,
the hypothetical possibilities if we con-
sider Clovis points, and similar points
found from coast to coast, to be repre-
sentative of a single people.
Age of Clovis Sites
Of six sites in the High Plains and
Southwest where Clovis artifacts have
been found in situ (6, 7), five have
now been dated by the radiocarbon
method. From the Lehner site in Ari-
zona (Fig. 1D) improved counting
techniques have clearly shown earlier
dates obtained by the solid-carbon
method to be in error (8). Statistical
treatment of six dates from the Clovis
level has demonstrated excellent agree-
ment between results for individual
samples and among separate labora-
tories and provides an average age of
11,260 ? 360 years (9). Overlying and
underlying sediments (samples A-33
bis and A-478b) have been dated
10,410 - 190 (10) and 11,600 ? 400
years ago, respectively (11).
The Dent site in Colorado (Fig.
1A), where Clovis points were first
found in situ and in association with
mammoth remains, has been dated
by the radiocarbon method mainly
through the efforts of George Agogino,
director of the Paleo-Indian Institute,
Eastern New Mexico University.
mammoth mandible from the original
excavation was made available for
analysis by H. M. Wormington, cura-
tor of archaeology at the Denver Mu-
seum of Natural History. The bone,
after being pulverized, was leached
with alcohol and then with acetone in
order to remove organic preservatives.
The progress of the leaching was mon-
itored through periodic analysis for ni-
trogen and organic carbon. The bone
was then pyrolyzed and leached with
hydrochloric acid to produce a residue
of "bone black" carbon. Bone from
which the preservatives had not been
leached (sample 1-473) was dated
7200 ? 200 years ago, while the
leached sample (1-622) was 11,200 -+
500 years ago.
As a result of intensive archeological
and geological activities during 1962-
63 at Blackwater No. 1 (Fig. IB),
the Clovis type site, our knowledge of
the Llano complex (6, 7) has been
substantially increased (12-14). These
activities produced materials suitable
for radiocarbon dating from both the
Clovis and the Folsom levels (9). Exca-
vation of the skeleton of a mammoth
produced carbonized plant remains in
a clay lens deposited in and around
the skull soon after the animal had
fallen prey to the Llano hunters. The
age of 11,170 - 360 years obtained for
sample A-481 dates the filling of the
mammoth's brain case with lacustro-
fluvial sediments and provides an early
date for the deposition of sediments
probably equivalent to the "brown
sand wedge" of Evans (see 6, p. 28).
Detailed geological investigations in
Blackwater Draw indicate the underly-
ing "gray sand" unit, in which many
of the Clovis artifacts and associated
mammoths occur, to be considerably
older than the cultural material (15)-
that is, the artifacts and at least some
of the animal remains have intruded
the gray sand.
A diatomite stratigraphic unit at
Blackwater No. 1 overlies the Clovis
material and contains Folsom artifacts
and associated bison remains. Carbon-
ized plants from this unit have pro-
vided dates of 10,490 - 900 (A-386)
and 10,250 ? 320 years (average of
A-379 and A-380) (9). These are com-
parable to dates for other Folsom
material.
At the Naco site in southeastern
Arizona (Fig. 1C) disseminated flecks
of charcoal collected from alluvium
containing a mammoth and associated
SCIENCE, VOL. 145
Fig. 1. Map showing (solid circles) radiocarbon-dated Clovis sites; (open circles) possible Clovis sites, radiocarbon-dated;
(squares) Clovis sites not dated; (triangles) sites in the Far North where fluted points have been discovered; and (grid pattern)
distribution of Clovis points found on the surface-all in relation to the approximate ice border during Two Creeks time and the
Bering land bridge when sea level was 55 meters lower than it is now. A, Dent site, Colorado; B, Blackwater No. 1 site, New
Mexico; C, Naco site, Arizona; D, Lehner site, Arizona; E, Bull Brook site, Massachusetts; F, Ventana Cave, Arizona; G, Union
Pacific Mammoth site, Wyoming; H, Domebo site, Oklahoma; I and J, Alaskan surface finds; K, Engigstciak site, Yukon Terri-
tory; L, Iyatayet site, Alaska; M, Shoop site, Pennsylvania; N, Williamson site, Virginia; 0, Quad site, Alabama; P, Miami site,
Texas; Q, Tonopah site, Nevada; R, Borax Lake site, California; S, Simon site, Idaho.
25 SEPTEMBER 1964 1409
Clovis points were dated 9250 ? 300 in Texas (21). Subsequently a date of graphic scrutiny. The Plainview, Texas,
years (A-9 + 10), but both the nature 10,780 ? 135 years was obtained for date of 9800 ? 500 years is for a
of the occurrence and the solid-carbon a sample (I-141) from the Folsom sample (L-303) which was not in
method of dating leave the date open level at the Lindenmeier site in Col- direct association with mammoth, and
to question (16). Geological evidence orado. This date permitted correlation the complex stratigraphy at the site
indicates that it is of the same age asof the geological stratigraphy at the makes the association tenuous at best.
Clovis material from the Lehner site. Lindenmeier site with that of other The Kassler, Colorado, date of 10,200
The Domebo site (Fig. 1H), near important sites of early man in the +- 350 years is for a mammoth tooth
Stecker, Oklahoma, is a recent discov- western United States (22). (W-401), a material which is notorious
ery of mammoth remains with associ- Recent investigations in Wyoming for contamination with more recent
ated fluted points that Anderson (17) have led to the discovery of two new materials. The Naco, Arizona, associa-
Folsom sites. The lowest cultural level
considers to be Clovis, with typological tion has been questioned, as mentioned
variations approaching those observed at the Brewster site in eastern Wyo- earlier. The one exception is Double
on some Plainview points. Wood sam- ming contains Folsom points in associ- Adobe, Arizona; there are now enough
ples (SM-610 and SM-695) from a ation with remains of Bison antiquus dates for that site to confirm the indi-
silt-clay unit containing the association
and is overlain by several horizons cated age of 8000 to 9000 years for
have been dated 1.0,123 ? 280 and containing Agate Basin points (23). A the Sulphur Springs deposits contain-
11,045 -- 647 years, respectively, the date of 10,375 + 700 years has been ing mammoth remains (10).
latter date being considered more per- obtained for the Folsom level (sample Other dates for mammoth remains-
tinent to the mammoth-man associa- 1-472), while two successive Agate dates obtained later than those cited
tion. Basin levels were dated 9900 ? 450 by Hester-are a minimum age of
One other dated site that may con- (M-1131) and 9350 + 450 years 11,900 ? 250 years (sample UCLA-
tain a Clovis occupation is Ventana (0-1252) (24). 637) for Tule Springs, Nevada (28),
Cave, Arizona (Fig. 1F); charcoal At the Hell Gap site near Guernsey, and an age of 11,280 ? 350 years for
(A-203) from the "volcanic debris" Wyoming, Folsom and Hell Gap arti- tusk (1-449) from the Union Pacific
unit (18, 19) at this site was dated facts occur in the same stratigraphic Mammoth site (see Fig. 1G) associ-
11,300 ? 1200 years (10). Dep- unit. While no date has as yet been ated with nondiagnostic artifacts (27).
osition of this unit was followed by a obtained for the Folsom level per se, The absence of mammoths from the
period of erosion, then by deposition a date of 10,850 ? 550 years obtained United States by the time the Folsom
which Bryan (see 18, p. 126) has at- for Hell Gap material (sample 1-167) and Hell Gap artifacts were deposited,
applies equally well to the Folsom
tributed to the Altithermal. It appears, 10,000 to 11,000 years ago, is evident
occupation, as determined from strati-
therefore, that the volcanic debris cov- in the stratigraphy of Paleo-Indian
ers a time span of several thousand graphic investigations (25). A loess sites from Montana to Arizona, with
years. A projectile point from this deposit below the Folsom and Hell the one exception of Double Adobe.
stratigraphic unit shows affinities with
Gap level, but above a clay unit dated Obviously, extinction of the mammoth
both Clovis and Folsom points; how- 13,060 ? 600 years (sample A-431), population did not occur overnight,
ever, Haury (19) points out that the is correlated with a similar deposit and it would not be surprising to find
age obtained by radiocarbon dating, near Rawhide Buttes, 50 kilometers mammoth remains as young as 10,500
the paucity of Folsom points in south- away, that contains mammoth remains years, reliably dated by radioisotope
eastern Arizona, and the small size of (A-366) older than 10,550 ? 350 and geological methods, but whether
some known Clovis points all indicate years (9). At Hartville, 24 kilometers any mammoth remains are as recent
the Ventana specimen to be Clovis. west of Hell Gap, a Clovis point wa- as 8000 to 9000 years ago must be
A Clovis site in the eastern United found in the same loess. considered an open question. It ap-
States, the Bull Brook site (Fig. 1E) From the radiocarbon evidence it pears, therefore, that the change from
in Massachusetts, has provided four appears, if the Naco date is omitted, Clovis points to Folsom points coin-
dates from scattered lumps of charcoal that the Folsom complex, representing cides with the relatively sudden dis-
(20). The dates range from 6940 - bison hunters, has a relatively short appearance of mammoths, and the
800 years (sample M-809) to 9300 ?- time span of 1000 years between concomitant switch to hunting Bison
400 years (M-807), but their applica- 10,000 and 11,000 years ago, and the antiquus in the High Plains suggests
bility to the artifacts has been ques- Llano complex, representing mammoth a change in the specialized hunting
tioned (1). hunters, a span of 500 years between weapon in response to a change in
11,000 and 11,500 years ago. The game.
transition from Clovis points to Fol-
Age of Folsom Sites som points occurred during a very
short period, there being few known Clovis Dispersion and Glaciation
Exclusive of caves, there are eight examples that might be considered
sites where Folsom artifacts have been transitional. The maximum suggested age for
found in situ. For five of these, in- A review of the pertinent dates led Clovis points does not appear to be
cluding Blackwater No. 1, dates have Hester (26) to conclude that mam- the result of premature interpretation
been obtained by the radiocarbon moths existed as late as 8000 years because geological investigation of the
method. ago in the United States, but none of known sites does not permit assump-
The earliest of these dates, 9883 ? the dates earlier than 11,000 years ago tion of a date for the Llano complex
350 years ago, was obtained by the are well established. With one excep- earlier than 12,000 years ago. At sites
solid-carbon method on burned bison tion all appear anomalous in the light where there are deposits immediately
bone (C-588) from the Lubbock site of later radiocarbon dating and strati- antedating the Clovis occupation these
1410 SCIENCE, VOL. 145
deposits indicate ecological conditions
as favorable to the existence of her-
bivorous megafauna, which man could
hunt, as conditions during the time of
the Clovis occupation, yet there is not
the slightest evidence of man's pres-
ence. This is not meant to imply that
there were no cultures in the New
World earlier than 12,000 years ago,
for there are good indications that
there were (29). It simply means that
we have no indisputable progenitor for
the Llano complex. Thus we find Clovis
points distributed over the southern
half of the continent around 11,300
years ago. Why this apparently sudden
appearance?
If we compare the geochronological
data with glacial events in the Lake
Michigan area, as recently reviewed by
Broecker and Farrand (30), we find a
striking relationship (Fig. 2). All dates
for Clovis artifacts, with the exception
Fluted Points in the Far North
Three fluted points have been found
on the surface in northwesternAlaska:
two in the western part of the Brooks
Range (Fig. 1J) and one, the Utukok
point, in the northwestern foothills of
the range (Fig. 11) (34, 35). The asso-
ciations are tenuous, but two poly-
hedral cores were found in the same
surface collections as two of the pro-
jectile points. All the points show mul-
tiple fluting and ground basal edges as
do most Clovis points; however, one of
CLOVIS
NORTH
AMER IC AN
41-~~~~%
GLAC /AT ION
the Alaskan specimens shows multiple
fluting to a more pronounced degree
than any Clovis point does. Three par-
allel flutes on each side of the Utukok
point bear a marked resemblance to
the flake scars left on a polyhedral
core by the removal of blades, suggest-
ing analogous techniques (5). In this
case the point is simply a core from
which blades are struck in the process
of fluting.
A fluted point was also found at the
Iyatayet site, Alaska (Fig. 1L), but
whether or not it is actually a part
FOLSOM
-N
0

of the Naco and Bull Brook dates, fall > o

41 L.
ai> L.
near or soon after the time of maxi- 0 ti a *^ ?
Q. 41 t. -*a
mum advance of ice during the Valders as a
0o
stade, and all the dates for Folsom ri ) -t -0 e
_c
S 0a
0
0C 0 -W 41 -J
artifacts are well within the period of 0
41: .
.-Q
4141 41 0 .^ L. -p
glacial recession. The seemingly abrupt t as 4* -J
>. i t Q) 03 Q ^t -i 03 o -J
appearance of Clovis sites in the High
Plains and the Southwest during the
Valders stade and after a period of
glacial recession suggests that a rela- 9O00 - -r
tionship exists between the appearance
of the Llano complex in the contermi- 0

nous United States and the Two Creeks 41

interstade, as considered by Hopkins u
-.
(31). The distribution of Clovis points 4A

in relation to dates, obtained by the

radiocarbon method, for ice and lake
borders in the Lake Michigan area
also suggests a Valders age, although
a late Mankato age is also a possibility
100o -

0
C.

bI0 t 4
I
(4, 32). The fact that Edmonton, Al- 3
berta, is the most northern locality in
which a Clovis point has been found I
K
(7, p. 42) is also pertinent. 11,000
From the glacial history of Canada I
and the Great Lakes area it now ap-
()
pears that by 12,000 years ago the
eastern foothills of the cordillera and
the western Great Plains of Canada VALDERS
(Fig. 1) were free of ice for the first
time in some 15,000 years (33). 12,o00 TWO
It is possible, therefore, that the
Clovis progenitors entered the United CREEKS
--'--'
States from Canada when an intergla- ?-

cier corridor opened up along the east-

ern foothills of the Canadian Rockies
MANKATO
during Two Creeks time. From the
seemingly rapid and wide dispersal of 13,000 -
Clovis points, or of very similar forms, -a

it appears that these people may have Fig. 2. Duration of Clovis and Folsom cultures, as revealed by radiocarbondating,
brought the technique of fluting with relativeto late periodsof glaciationin North America.[Glaciationdata from Broecker
them. and Farrand(30)]
25 SEPTEMBER 1964 1411
of the Denbigh Flint Complex, dated
4500 to 5000 years ago, is open to
question, as it is not uncommon for an
artifact assemblage to contain anoma-
lous specimens from an earlier com-
plex (36). Crude fluted points from the
Engigstciak site (Fig. 1K) near the
Yukon arctic coast are believed to have
been deposited soon after the last gla-
ciation of the area (37). If this corre-
sponds to the last glaciation recorded
in the northeastern Mackenzie Delta,
which ended before 12,000 years ago
(38), the artifacts could be 12,000 to
15,000 years old.
In considering the origin of fluted-
point cultures, the material from the
Far North, while scanty and incon-
clusive, cannot be ignored. From a
typological point of view the Utukok
fluted point is more like Clovis than
like Folsom points; the general shape,
the percussion flaking, and the removal
of three blades in fluting are all traits
more akin to Clovis than to Folsom
points. If the Utukok point differs
from Clovis in age, it is probably older,
as suggested by Witthoft (2), rather
than younger; if it is older, the absence
of fluted points from Siberia would
point to the Bering land bridge, Alaska,
or northern Canada as a possible place
of origin for fluting.
Green (13) has suggested that a re-
cent discovery of blades believed to be
of Clovis age at Blackwater No. 1,
New Mexico, may be the first tangible
link in cultural tradition between the
New World Paleo-Indian and the Up-
per Paleolithic blade industries of the
Old World. The blades, polyhedral
cores, and multifluted points of the
Enterline Chert Industry may be quite
as good a link as the Clovis blades
(2), but, as emphasized by Wilmsen
(39), the Alaskan and Canadian ma-
terial supports the view that there were
ties between the New World Paleo-
Indian and the Siberian Upper Paleo-
lithic blade industries.
In considering the distribution of
Clovis points and the Alaskan and
Yukon fluted-point finds in relation to
late glacial geochronology we find a
logical sequence of events, and the
pieces begin to fall into place. If
Clovis progenitors traversed a corridor
through Canada around 12,000 years
ago and dispersedthroughout the Unit-
ed States south of the Valders ice bor-
der in the ensuing 700 years, then they
probably were in Alaska some 500
years earlier-say, 12,500 years ago.
The Alaskan fluted points and some
of the polyhedral cores could represent
1412
this occupation and could, therefore,
be ancestral to Clovis points and
blades.
Both the Arctic slope and central
Alaska were essentially ice-free during
late Wisconsin time and could have
been migration routes, but the known
finds of fluted points suggest the
northern route for Clovis progenitors.
Furthermore, a tundra-covered area
between eastern Siberia and western
Alaska around 13,000 years ago, when
sea level was 55 to 60 meters lower
than it is today, would have presented
no obstacle to people already accli-
mated to a tundra environment. Such
an environmentappears to have existed
at Nome, across the arctic slope, and
in the Canadian Great Plains during
the recession of Wisconsin ice (35,
40).
The dispersion of Clovis peoples in
response to opening of a glacial bar-
rier might explain the abrupt increase,
at 10,400 years ago, in the number of
dates shown in Libby's histogram of
archeological dates obtained by the
radiocarbon method (41). Five Clovis
dates, extending the point of abrupt
increase back to 11,300 years ago, can
now be added. From at least. 25,000
to 12,500 years ago, the trans-Canadian
migration route was blocked by co-
alescence of the Cordilleran and Laur-
entide ice sheets. Therefore, any pre-
Clovis archeological remains south of
Canada probably derive from migra-
tions earlier than "classical"Wisconsin
time. The paucity of radiocarbon-dated
sites older than 11,500 years suggests a the United States around 11,500 years
sparse population at best.
According to the schedule proposed
here, the dispersal of people ancestral
to Clovis man from Siberia to the
southwestern United States would have
taken 1500 years and would represent
a unique "migration,"as suggested by
Green (13). Once south of the ice bor-
der, Llano man dispersed from Cali-
fornia to Massachusettsand southward
into Mexico (6), in less than 1000
years. The subsequent spread of early
cultures throughout the New World
between 11,000 and 10,000 years ago
attests to the rapidity with which ab-
original bands could disperse in the
pursuit of big game-first mammoths
and then bison.
Summary and Conclusions
The stratigraphicrecord shows Clo- elsewhere. Clovis progenitors might
vis projectile points to be restricted to best be sought in northern Alaska or
sediments between 11,000 and 11,500 the MackenzieValley.
years old. Underlying deposits dating
back 11,600 to 13,000 years are with-
out evidence of human occupation. In
the High Plains, overlying deposits
dating back 10,000 to 11,000 years
contain Folsom and Hell Gap artifacts
and are without mammoth remains.
The glacial history of Alaska, Can-
ada, and the Great Lakes region indi-
cates that, for the first time in at least
15,000 years, an ice-free, trans-Cana-
dian corridor opened up approximately
12,000 years ago. Since Clovis points
are distributed from coast to coast
south of the Valders ice border, the
abrupt appearance of Clovis artifacts
in the stratigraphicrecord of the High
Plains some 700 years later suggests
that Clovis progenitors passed through
Canada during Two Creeks time. If
eastern fluted points (for example, En-
terline) are older than Clovis points,
the difference may be on the order of
only a hundred or so years, not
thousands.
The change from Clovis points to
Folsom points in the High Plains may
be related to a marked decline in the
mammoth population after 11,000
years ago, but whether or not man
was a prime factor in the extinction
of Pleistocene megafauna is a moot
question.
On the basis of new data and criti-
cal geological evaluation of dates ob-
tained by the radiocarbon method a
hypothesis has been offered to explain
(i) the abrupt appearance of Clovis
points in the stratigraphic record of
ago, and (ii) the lack of a cultural
continuum in the United States leading
to fluted projectile points. Llano hunt-
ers, like the game they pursued, may
have persisted longer in some areas
of the continent (for example, Bull
Brook) than in others, but if a Clovis
site can be found for which good strati-
graphic evidence supports a date ear-
lier than the Two Creeks interstade,
then correlation of this event to the
opening of the trans-Canadianice-free
corridor is incorrect (see 41a). Such a
misinterpretationof timing would not
affect the explanation for the lack of
Clovis progenitors in the United States.
We must continue to look for an in-
digenous cultural continuum leading to
Clovis points, but if such cannot be
demonstratedin the conterminousUnit-
ed States, then it would appear that
fluted projectile points were developed
SCIENCE, VOL. 145
 The interpretations offered here are
based on new data and critical geo-
logical evaluation of dates previously
obtained by the radiocarbon method.
How valid these interpretations are
can be ascertained only through care-
ful scrutiny of all man-mammoth asso-
ciations found in the future, to assure
precise relating of dates, fossils, and
artifacts to the stratigraphic frame-
work. We must pay closer attention
to stratigraphic detail if we are to
make the fullest use of radiocarbon
dating.
References and Notes
1. R. J. Mason, Current Anthropol. 3, 227
(1962).
2. J. Witthoft, Am. Antiquity 19, 271 (1954);
, Current Anthropol. 3, 267 (1962);
W. B. Roosa, ibid., p. 263; --- , Mich.
Archaeologist 9, 44 (1963).
3. A. D. Krieger, Am. Antiquity 19, 273 (1954).
4. R. J. Mason, Univ. Mich. Anthropol. Papers
11, 12 (1958).
5. J. Witthoft, Proc. Am. Phil. Soc. 96, 464
(1952).
6. E. H. Sellards, Early Man in America (Univ.
of Texas Press, Austin, 1952).
7. H. M. Wormington, "Ancient Man in North
America," Denver Museum Nat. Hist. Publ.
(1957).
8. E. W. Haury, E. B. Sayles, W. W. Wasley,
Am. Antiquity 25, 2 (1959).
9. P. E. Damon, C. V. Haynes, A. Long,
Radiocarbon 6, 100 (1964).
10. P. E. Damon and A. Long, ibid. 4, 239
(1962).
11. Unpublished observations.
12. J. J. Hester and J. M. Warnica, paper pre-
sented at the 28th annual meeting of the
Society for American Archaeology, Boulder,
1963.
13. F. E. Green, Am. Antiquity 29, 145 (1963).
14. Eastern New Mexico University's 1963 ex-
cavation project was directed by George
Agogino.
15. V. Haynes, "Pleistocene and Recent Stratig-
raphy of Blackwater Draw, New Mexico and
Rich Lake, Texas," in press.
16. E. W. Haury, Am. Antiquity 19, 1 (1953);
E. N. Wise and D. Shutler, Jr., Science 127,
72 (1958); E. Antevs, Am. Antiquity 25,
31 (1959).
17. A. D. Anderson, News Bull. Great Plains
Hist. Assn. 2, 4 (1962). More recent data
are included with the kind permission of
Adrian Anderson, Marvin Tong, Frank
Leonhardy, and Claude Albritton, who are
preparing a monograph on the interdisci-
plinary investigations of the Domebo site.
18. E. W. Haury, Ventana Cave (Univ. of New
Mexico Press, Albuquerque, 1950).
19. - , personal communication.
20. D. S. Beyers, Am. Antiquity 24, 427 (1959).
21. W. F. Libby, Radiocarbon Dating (Univ. of
Chicago Press, Chicago, 1955), p. 107.
22. V. Haynes and G. A. Agogino, Proc. Den-
ver Museum Nat. Hist. 9, 23 (1960).
23. G. A. Agogino and W. D. Frankforter,
Master Key 34, 102 (1960); F. H. H. Rob-
erts, "The Agate Basin Complex," in Home-
naje a Pablo Martinez del Rio en el xxv
aniversario de la edicion de los origenes
Americanos (Mexico, 1961).
24. H. R. Crane and J. B. Griffin, Radiocarbon
5, 244 (1962).
25. H. T. Irwin, G. A. Agogino, C. C. Irwin,
paper presented at the 27th annual meeting of
the Society for American Archaeology, Tuc-
son (1962); V. Haynes, "Late Pleistocene
and Recent Stratigraphy of the Hell Gap
Area," in preparation.
26. J. J. Hester, Am. Antiquity 26, 58 (1960).
27. G. J. Fergusson and W. F. Libby, Radio-
carbon 6, 321 (1964).
28. C. Irwin, H. Irwin, G. A. Agogino, Natl.
Geographic Mag. 121, 828 (1962).
29. A. D. Krieger, Am. Antiquity 28, 138 (1962);
H. M. Wormington, Am. Scientist 50, 230
(1962).
30. W. S. Broecker and W. R. Farrand, Bull.
Geol. Soc. Am. 74, 795 (1963).
31. D. M. Hopkins, Current Anthropol. 3, 254
(1962).
32. G. I. Quimby, Am. Antiquity 28, 558 (1963);
Current Anthropol. 3, 262 (1962).
33. R. F. Flint, Geol. Soc. Am. Spec. Paper 60
(1945), pt. 1; L. Horberg, Bull. Geol. Soc.
Am. 65, 1093 (1954); B. G. Craig and J. G.
Fyles, Geol. Survey. Can. Paper 60-10 (1960);
J. L. Hough, Am. Scientist 51, 84 (1963);
J. A. Elson, Science 126, 999 (1957); B. G.
Craig, Geol. Soc. Am. Abstr., in press;
, personal communication.
34. R. S. Solecki, Am. Antiquity 17, 55 (1950);
R. M. Thompson, ibid. 14, 62 (1948).
35. R. S. Solecki, Smithsonian Inst. Ann. Rept.
1950, 469 (1951).
36. J. L. Giddings, Am. Antiquity 16, 193 (1951);
H. B. Collins, ibid. 18, 199 (1953).
37. R. S. MacNeish, Univ. Alaska Anthropol.
Papers 4, 91 (1956); J. R. Mackay, W. H.
Mathews, R. S. MacNeish, Arctic 14, 25
(1961).
38. F. Muller, Arctic 15, 279 (1962).
39. E. N. Wilmsen, Am. Antiquity 29, 338
(1964).
40. T. N. V. Karlstrom, Ann. N.Y. Acad. Scl.
95, 290 (1961); D. M. Hopkins, Science 129,
1519 (1959); , F. S. Macneil, E. B.
Leopold, Intern. Geol. Congr. 21st, Copen-
hagen, 1960, Rept. Session, Norden (1960),
pt. 4, p. 46; W. 0. Kupsch, Am. J. Sci.
258, 282 (1960); S. C. Porter, Geol. Soc.
Am. Abstr. 1962, 216 (1963).
41. W. F. Libby, Science 133, 621 (1961).
41a. My attention has been called to a pertinent
paper by 0. H. Prufer and R. S. Baby
[Paleo-Indians of Ohio (Ohio Historical
Society, Columbus, 1963)]. These authors in-
terpret the distribution of some fluted points
in relation to glacial moraines and ancient
beach ridges as evidence for occupation of
Ohio by makers of fluted points before the
Two Creeks interstade. The geochronology of
fluted points in the East has been plagued
by the almost total lack of stratigraphic
context that would permit definitive estima-
tion of age. In the absence of such context,
much reliance has necessarily been placed
upon the surface distribution of fluted points
in relation to datable geomorphic features,
but as Prufer and Baby state (p. 53),
"mere association with a certain beach sys-
tem [or moraine] does not preclude the pos-
sibility that a specimen so discovered post-
dates the formation of this particular system;
such association merely implies a maximum
date." It should be added that the temporal
position of the geomorphic features themselves
is not adequately known. For the moment
it appears that fluted points occupy the base
of projectile-point sequences, but the de-
finitive age of fluted points in the East must
await the finding of more sites with strati-
graphically controlled dates and artifacts.
42. This article is contribution No. 89 of the
University of Arizona's Program in Geochro-
nology. The radiocarbon dating and field
investigations were in part supported by the
National Science Foundation (grant GP
2330), the Research Corporation (unrestricted
venture grant), the American Philosophical
Society, and the National Geographic Society.
Much-appreciated comments on the manu-
script were provided by P. E. Damon, E. W.
Haury, T. L. Smiley, H. M. Wormington,
D. M. Hopkins, and B. G. Craig. I bear all
responsibility for errors, omissions, and mis-
interpretations.

attributes of science and of the crack-

pot to help in making this decision.
Let me illustrate what I mean by
the failure of any single test. A scien-
A Measure for Crackpots tist generally strengthens his stand
greatly by his ability to predict. Ein-
stein's early work in the general
How does one distinguish between valid theory of relativity gained credence
scientific work and counterfeit "science"? by accounting in part for a known
error in the perihelion of the planet
Mercury. His theories gained real ac-
Fred J. Gruenberger ceptance, however, many years later
when British astronomers (during
World War I, when Einstein was a
citizen of an enemy country) verified
For every article one sees in a techni- infallibly between crackpot work and by direct observation his statements
cal journal or, for that matter, even good science. For every criterion that about the previously unsuspected bend-
in the public press, a decision has to is advanced one can quickly think of a ing of light in a gravitational field.
be made: Is this worth reading or is counter example. Every idea that is of- Here was the principle of predictability
it something that can safely be fered to us falls somewhere between used to the fullest.
If predictability is to be used as a
skipped? It would seem that there is the two extremes. Since there is no
no acid test or even a small group single test, the best that can be done
The author is affiliated with the Rand Corpora-
of tests which will serve to distinguish is to offer a checklist of some of the tion, Santa Monica, Calif.
25 SEPTEMBER 1964 1413